Spatial-temporal multi-scale interaction for few-shot video summarization

Qun Li, Zhuxi Zhan, Yanchao Li, Bir Bhanu

Abstract: Highlights•We propose a new task, called few-shot video summarization.•We propose a baseline for few-shot video summarization, namely STeMI.•STeMI endeavors to extract spatial-temporal representations by using a dual-branch.•We adapt several SOTAs for video summarization to few-shot scenarios.•We conduct comparative evaluations against our proposed STeMI model.